Date: Mon, 20 Apr 1998 21:27:48 -0700 From: Erik Viirre X-Mailer: Mozilla 3.0 (Macintosh; I; PPC) MIME-Version: 1.0 To: Edward Lipson Subject: Re: Fwd: Important Information about KDI Grant Proposals Ed, Here is some stuff to start. Sorry for the delay. Please page me if you have any questions (800 573-4775). I'll add some more as we go. Let me know what you think. Does this fit in? III. Video Oculography (VOG)     The VOG Project will develop an eye movement test battery for the learning environment consisting of eye movement tracking system embedded in lightweight goggles that subjects can wear in the real world, while viewing a semi-immersive system or while they are immersed in VR with an HMD. The goggles will have integrated video-cameras that image the eye and an image processing computer will monitor the subject’d gaze direction. The project plan includes three phases: integration of equipment and refinement of test statistics, Eye movement pattern study in VR application development environment and testing of subjects with Eye Movement Test battery in the actual learning environment.     Dr. Viirre has extensive experience in eye movement research, construction of eye movement tracking hardware and integration of the software. In year 1, he will construct lightweight goggles that subjects can wear in the real world or while viewing a semi-immersive  system attached CrystalEyes. The goggles will have integrated Video-oculography systems (VOG) that will image the eye. The images will be analyzed by a dedicated eye tracking computer built by ISCAN Corporation. He will also construct a head mounted display (HMD) with light weight cameras that will eye track using the same computer. The eye tracking computer will be connected by high speed LAN to a data analysis computer that will also have the epoch markers from the main VR system. An integrated analysis program developed in the Lab View environment (National Instruments Corp.) will produce statistics on eye movement pattern epochs and their correlation with other activities in the virtual environment. The statistical tools will be similar to “Maceyeball” the eye movement analysis package developed at UCLA. The development time for this new system will be one year. The eye trackers will be used both in the PP and VR labs. Task Analysis Year 1 VOG1. In this task, the existing eye tracking system will be used as a baseline system. Margaret Murray (App), John Hestenes (PHY) and Erik Viirre (VOG) will establish the baseline software requirements for the eye pattern tool. This will include: eyetrack parameters, event markers, and statistical analysis. These tools will be integrated into the APP and PHY protocols. The baseline eyetracking system will only measure relative eye position, as position calibration is unstable over long periods. Eye movement patterns will be detectable. VOG2. Based on the requirements outlined in VOG1, the baseline tool will be created. Identifiers for reading, 3D eye movements (variation in depth),  and tracking versus head stabilization movements will be made. VOG3. Using the content neutral application from the APP project, baseline studies will be carried out with the basic eye tracking tool. These studies will test approximately 6 subjects in brief sessions. There will be verification of the computerized task analysis with video tape review of the session. VOG4. There will be a review of the analysis protocols and results of the initial tests by the APP team. VOG5. Based on the review of the first year’s results, there will be a setting of the second generation of the eye tracking pattern analysis system. The new system will incorporate the new eye tracking hardware. This hardware will allow identification of gaze direction and therefore object of regard. Thus new components within the APP will allow object identification with direction of regard to be event marked. This will be a new capability allowing the gaze dwell time on objects to be determined. VOG6. Users in the research lesson developed for the second year will be analyzed using the baseline eye pattern detector. VOG7. In the second year, the new software will be developed and integrated with the new eyetracking mechanism. VOG8. There will be testing and analysis of the second generation pattern detector  in research lesson #2. Only one setup of the research group will be instrumented. VOG9. With the help of the APP and research lesson teams, the eye movement pattern data will be reviewed and summarized. VOG10. The eye tracking of 2 users will be analyzed in a few test sessions. VOG11. The eye pattern analysis techniques, data and conclusions will be prepared for publication. -- Erik Viirre M.D. Ph.D. Research Scientist Human Interface Technology Laboratory University of Washington Box 352142       "Man is still the most Extraordinary Computer of all" Seattle WA 98107-2142                     JFK (206) 616-3071 fax    543-5380 http://www.hitl.washington.edu